Dynamic monitoring of chemical processes based on latent variable auto-regressive algorithm

doi:10.11949/j.issn.0438-1157.20181180

Abstract

Abstract:

From the viewpoint of constructing auto-regressive (AR) models for latent variables, the current paper proposed a dynamic modeling and monitoring approach for chemical processes based on latent variable auto-regressive (LVAR) algorithm. With respect to the requirement of minimizing AR model residual, the LVAR algorithm simultaneously searches for projecting vectors and AR coefficient vectors, so as to extracting dynamic latent variables and their corresponding AR models. In addition, the LVAR algorithm can efficiently distinguish the auto-correlated and cross-correlated relations inherited in the sampled data, by extracting the dynamic latent variables first and then the static component information. In the comparisons, the superiority and validity of the LVAR method are demonstrated by comparing the fault monitoring results in a classical chemical plant with other three counterparts.

Key words: principal component analysis, fault detection, missing data, process systems

摘要：

从建立潜变量自回归（AR）模型的角度出发，提出了一种基于潜变量自回归（LVAR）算法的化工过程动态建模与监测方法，旨在提取动态潜变量的同时给出各潜变量的AR模型。LVAR算法在最小化潜变量的AR模型残差的约束下，通过同时搜寻投影变换向量与AR系数向量，实现了对动态潜变量的特征提取及其AR模型的建立。此外，LVAR算法通过先提取动态潜变量后提取静态成分信息的方式，有效地区分了采样数据中的自相关性与交叉相关性。在对比实验中，通过比较分析LVAR方法与其他三种典型的动态过程监测方法在经典化工过程对象上的故障监测结果，验证了LVAR方法在动态过程监测上的优越性与可靠性。

关键词: 主成分分析, 故障检测, 缺失数据, 过程系统

CLC Number:

TP 277

Junmiao TANG, Haizhen YU, Xuhua SHI, Chudong TONG. Dynamic monitoring of chemical processes based on latent variable auto-regressive algorithm[J]. CIESC Journal, 2019, 70(3): 987-994.

唐俊苗, 俞海珍, 史旭华, 童楚东. 基于潜变量自回归算法的化工过程动态监测方法[J]. 化工学报, 2019, 70(3): 987-994.

Figures/Tables 5

Fig.1 Auto-correlation plots of dynamic latent variable

Fig.2 Auto-correlation plots of column vectors in residual matrix FD+1

Fig.3 Comparison of average missing alarm rates

Table 1 False alarm rates and missing alarm rates achieved in TE process

No.	DPCA		DLV			DiPCA			LVAR
No.	$T 2$	$Q$	$T a 2$	$T s 2$	$Q s$	$T a 2$	$T s 2$	$Q s$	$T a 2$	$T s 2$
1	0.75	0.13	0	0	0.63	0.25	0.88	0.63	0	0.88
2	1.50	2.75	1.00	2.25	9.63	1.63	1.88	3.50	1.63	4.13
4	94.00	0	95.85	0	87.13	0	77.25	0	0.13	0
5	76.00	55.00	89.95	81.75	0.13	76.69	93.63	79.38	0	0
6	1.13	0	0.25	1.00	0	0.50	1.63	0	0	0
7	0	0	82.28	0	69.75	0	0	0	58.50	0
8	2.63	3.00	12.19	6.38	30.50	5.14	16.88	19.50	1.88	25.63
10	75.13	48.88	94.22	90.50	37.50	83.33	97.25	71.13	19.63	39.25
11	72.88	6.00	89.32	19.00	88.50	14.79	58.25	12.63	23.50	39.88
12	0.88	3.63	19.72	9.00	17.38	8.15	21.00	9.88	0.25	3.13
13	5.75	4.63	29.27	4.88	11.00	6.52	26.63	7.75	4.50	7.63
14	0.13	0	0	0	11.25	0	12.75	0	0	24.13
16	90.38	48.00	91.08	90.63	36.63	81.70	98.38	69.50	27.38	16.88
17	22.75	2.25	31.28	5.13	6.25	11.53	33.13	3.25	2.25	12.75
18	11.13	9.38	11.43	10.00	9.63	9.52	10.88	9.50	9.88	9.63
19	77.25	33.38	90.58	81.13	37.00	60.53	82.25	72.63	41.88	11.63
20	61.13	36.38	73.12	60.63	35.13	50.63	78.75	47.38	47.63	25.13
21	54.25	49.50	92.46	49.25	83.13	73.56	78.13	74.63	71.38	68.5
Ave.	35.98	16.83	50.22	28.42	31.73	26.92	43.86	26.74	16.54	15.28

Table 1 False alarm rates and missing alarm rates achieved in TE process

No.	DPCA		DLV			DiPCA			LVAR
No.	$T 2$	$Q$	$T a 2$	$T s 2$	$Q s$	$T a 2$	$T s 2$	$Q s$	$T a 2$	$T s 2$
1	0.75	0.13	0	0	0.63	0.25	0.88	0.63	0	0.88
2	1.50	2.75	1.00	2.25	9.63	1.63	1.88	3.50	1.63	4.13
4	94.00	0	95.85	0	87.13	0	77.25	0	0.13	0
5	76.00	55.00	89.95	81.75	0.13	76.69	93.63	79.38	0	0
6	1.13	0	0.25	1.00	0	0.50	1.63	0	0	0
7	0	0	82.28	0	69.75	0	0	0	58.50	0
8	2.63	3.00	12.19	6.38	30.50	5.14	16.88	19.50	1.88	25.63
10	75.13	48.88	94.22	90.50	37.50	83.33	97.25	71.13	19.63	39.25
11	72.88	6.00	89.32	19.00	88.50	14.79	58.25	12.63	23.50	39.88
12	0.88	3.63	19.72	9.00	17.38	8.15	21.00	9.88	0.25	3.13
13	5.75	4.63	29.27	4.88	11.00	6.52	26.63	7.75	4.50	7.63
14	0.13	0	0	0	11.25	0	12.75	0	0	24.13
16	90.38	48.00	91.08	90.63	36.63	81.70	98.38	69.50	27.38	16.88
17	22.75	2.25	31.28	5.13	6.25	11.53	33.13	3.25	2.25	12.75
18	11.13	9.38	11.43	10.00	9.63	9.52	10.88	9.50	9.88	9.63
19	77.25	33.38	90.58	81.13	37.00	60.53	82.25	72.63	41.88	11.63
20	61.13	36.38	73.12	60.63	35.13	50.63	78.75	47.38	47.63	25.13
21	54.25	49.50	92.46	49.25	83.13	73.56	78.13	74.63	71.38	68.5
Ave.	35.98	16.83	50.22	28.42	31.73	26.92	43.86	26.74	16.54	15.28

Fig.4 Monitoring charts of fault 10

References 32

1	TidririK, ChattinN, VerronS, et al. Bridging data-driven and model-based approaches for process fault diagnosis and health monitoring: a review of researches and future challenges [J]. Annu. Rev. Control, 2016, 42: 63-81.
2	TongC, El-FarraN H, PalazogluA, et al. Fault detection and isolation in hybrid process systems using a combined data-driven and observer-design methodology [J]. AIChE J., 2014, 60(8): 2805-2814.
3	ZhouL, ZhengJ, GeZ, et al. Multimode process monitoring based on switching autoregressive dynamic latent variable model [J]. IEEE Trans. Ind. Electron., 2018, 65(10): 8184-8194.
4	GeZ. Review on data-driven modeling and monitoring for plant-wide industrial processes [J]. Chemom. Intell. Lab. Syst., 2017, 171: 16-25.
5	SeversonK, ChaiwatanodomP, BraatzR D. Perspectives on process monitoring of industrial systems [J]. Annu. Rev. Control, 2016, 42: 190-200.
6	DongY, QinS J. Regression on dynamic PLS structures for supervised learning of dynamic data [J]. J. Process Control, 2018, 68: 64-72.
7	蒋栋年, 李炜. 基于数据驱动残差评价策略的故障检测方法[J]. 控制与决策, 2017, 32(7): 1181-1188.
	JiangD N, LiW. Fault detection method based on data-driven residual evaluation strategy [J]. Control & Decision, 2017, 32(7): 1181-1188.
8	GeZ, SongZ, DingS, et al. Data mining and analytics in process industry: the role of machine learning [J]. IEEE Access, 2017, 5: 20590-20616.
9	董顺, 李益国, 孙栓柱, 等. 基于状态空间主成分分析网络的故障检测方法[J]. 化工学报, 2018, 69(8): 3528-3536.
	DongS, LiY G, SunS Z, et al. Fault detection method based on state space-PCANet [J]. CIESC Journal, 2018, 69(8): 3528-3536.
10	童楚东, 蓝艇, 史旭华. 基于互信息的分散式动态PCA故障检测方法[J]. 化工学报, 2016, 67(10): 4317-4323.
	TongC D, LanT, ShiX H. Fault detection by decentralized dynamic PCA algorithm on mutual information [J]. CIESC Journal, 2016, 67(10): 4317-4323.
11	赵帅, 宋冰, 侍洪波. 基于加权互信息主元分析算法的质量相关故障检测[J]. 化工学报, 2018, 69(3): 926-973.
	ZhaoS, SongB, ShiH B. Quality-related fault detection based on weighted mutual information principal component analysis [J]. CIESC Journal, 2018, 69(3): 926-973.
12	LiuY, ZhangG, XuB. Compressive sparse principal component analysis for process supervisory monitoring and fault detection [J]. J. Process Control, 2017, 50: 1-10.
13	蓝艇, 童楚东, 史旭华. 变量加权型主元分析算法及其在故障检测中的应用[J]. 化工学报, 2017, 68(8): 3177-3182.
	LanT, TongC D, ShiX H. Variable weighted principal component analysis algorithm and its application in fault detection [J]. CIESC Journal, 2017, 68(8): 3177-3182.
14	TongC, LanT, ShiX. Fault detection and diagnosis of dynamic processes using weighted dynamic decentralized PCA approach [J]. Chemom. Intell. Lab. Syst., 2017, 161: 34-42.
15	XiaoZ, WangH, ZhouJ. Robust dynamic process monitoring based on sparse representation preserving embedding [J]. J. Process Control, 2016, 40: 119-133.
16	LeeJ M, YooC K, LeeI B. Statistical monitoring of dynamic processes based on dynamic independent component analysis [J]. Chem. Eng. Sci., 2004, 59: 2995-3006.
17	ZhaoH. Dynamic graph embedding for fault detection [J]. Comput. Chem. Eng., 2018, 117: 359-371.
18	Vanden KerkhofP, GinsG, VanlaerJ, et al. Dynamic model-based fault diagnosis for (bio)chemical batch processes [J]. Comput. Chem. Eng., 2012, 40: 12-21.
19	ChoiS W, MorrisJ, LeeI B. Dynamic model-based batch process monitoring [J]. Chem. Eng. Sci., 2008, 63: 622-636.
20	ZhaoC, GaoF. Online fault prognosis with relative deviation analysis and vector autoregressive modeling [J]. Chem. Eng. Sci., 2015, 138: 531-543.
21	LiG, QinS J, ZhouD. A new method of dynamic latent-variable modeling for process monitoring [J]. IEEE Trans. Ind. Electron., 2014, 61(11): 6438-6445.
22	DongY, QinS J. A novel dynamic PCA algorithm for dynamic data modeling and process monitoring [J]. J. Process Control, 2018, 67: 1-11.
23	ZhouL, LiG, SongZ, QinS J. Autoregressvie dynamic latent variable models for process monitoring [J]. IEEE Tran. Control Syst. Tech., 2017, 25(1): 366-373.
24	ShangJ, ChenM, ZhangH. Fault detection based on augmented kernel Mahalanobis distance for nonlinear dynamic processes[J]. Comput. Chem. Eng., 2018, 109: 311-321.
25	DownsJ J, VogekE F. A plant-wide industrial process control problem [J]. Comput. Chem. Eng., 1993, 17(3): 245-255.
26	仓文涛, 杨慧中. 基于主元子空间富信息重构的过程监测方法[J]. 化工学报, 2018, 69(3): 1114-1120.
	CangW T, YangH Z. A process monitoring method based on informative principal component subspace reconstruction [J]. CIESC Journal, 2018, 69(3): 1114-1120.
27	周乐, 宋执环, 侯北平, 等. 一种鲁棒半监督建模方法及其在化工过程故障检测中的应用[J]. 化工学报, 2017, 68(3): 1109-1115.
	ZhouL, SongZ H, HouB P, et al. Robust semi-supervised modelling method and its application to fault detection in chemical processes [J]. CIESC Journal, 2017, 68(3): 1109-1115.
28	QinS J. Statistical process monitoring: basics and beyond [J]. J. Chemom., 2003, 17(7/8): 480-502.
29	杨健, 宋冰, 谭帅, 等. 时序约束NPE算法在化工过程故障检测中的应用[J]. 化工学报, 2016, 67(12): 5131-5139.
	YangJ, SongB, TanS, et al. Time constrained NPE for fault detection in chemical processes [J]. CIESC Journal, 2016, 67(12): 5131-5139.
30	BakdiA, KouadriA. A new adaptive PCA based thresholding scheme for fault detection in complex systems [J]. Chemom. Intell. Lab. Syst., 2017, 162: 83-93.
31	孙栓柱, 董顺, 江叶峰, 等. 基于最小充分统计量模式分析的故障检测方法[J]. 化工学报, 2018, 69(3): 1228-1237.
	SunS Z, DongS F, JiangY F, et al. Fault detection method based on minimum sufficient statistics pattern analysis [J]. CIESC Journal, 2018, 69(3): 1228-1237.
32	LiuK, FeiZ, YueB, et al. Adaptive sparse principal component analysis for enhanced process monitoring and fault isolation [J]. Chemom. Intell. Lab. Syst., 2015, 146: 426-436.

[1]	Yue CAO, Chong YU, Zhi LI, Minglei YANG. Industrial data driven transition state detection with multi-mode switching of a hydrocracking unit [J]. CIESC Journal, 2023, 74(9): 3841-3854.
[2]	Zhewen CHEN, Junjie WEI, Yuming ZHANG. System integration and energy conversion mechanism of the power technology with integrated supercritical water gasification of coal and SOFC [J]. CIESC Journal, 2023, 74(9): 3888-3902.
[3]	Yihao ZHANG, Zhenlei WANG. Fault detection using grouped support vector data description based on maximum information coefficient [J]. CIESC Journal, 2023, 74(9): 3865-3878.
[4]	Yuyuan ZHENG, Zhiwei GE, Xiangyu HAN, Liang WANG, Haisheng CHEN. Progress and prospect of medium and high temperature thermochemical energy storage of calcium-based materials [J]. CIESC Journal, 2023, 74(8): 3171-3192.
[5]	Guang WANG, Fashun SHAN, Yucheng QIAN, Jianfang JIAO. Incipient fault detection method for chemical process based on ensemble learning transfer entropy [J]. CIESC Journal, 2023, 74(7): 2967-2978.
[6]	Guixian LI, Abo CAO, Wenliang MENG, Dongliang WANG, Yong YANG, Huairong ZHOU. Process design and evaluation of CO₂ to methanol coupled with SOEC [J]. CIESC Journal, 2023, 74(7): 2999-3009.
[7]	Yuanzhe SHAO, Zhonggai ZHAO, Fei LIU. Quality-related non-stationary process fault detection method by common trends model [J]. CIESC Journal, 2023, 74(6): 2522-2537.
[8]	Cheng YUN, Qianlin WANG, Feng CHEN, Xin ZHANG, Zhan DOU, Tingjun YAN. Deep-mining risk evolution path of chemical processes based on community structure [J]. CIESC Journal, 2023, 74(4): 1639-1650.
[9]	Bing SONG, Chengfeng ZHENG, Hongbo SHI, Yang TAO, Shuai TAN. Research on quality-related fault detection method based on VAE-OCCA [J]. CIESC Journal, 2023, 74(4): 1630-1638.
[10]	Minghui YANG, Xiaoyue LIU, Xiaogang DENG, Mingyan LIAO, Chunwang HOU. Incipient fault detection for dynamic chemical processes based on weighted probability CVDA [J]. CIESC Journal, 2022, 73(9): 3963-3972.
[11]	Yalin WANG, Yuqing PAN, Chenliang LIU. Intermittent process monitoring based on GSA-LSTM dynamic structure feature extraction [J]. CIESC Journal, 2022, 73(9): 3994-4002.
[12]	Jiawang YONG, Qianqian ZHAO, Nenglian FENG. Fault diagnosis of proton exchange membrane fuel cell based on nonlinear dynamic model [J]. CIESC Journal, 2022, 73(9): 3983-3993.
[13]	Jinyu GUO, Zhe WANG, Yuan LI. Fault detection method based on kernel entropy independent component analysis [J]. CIESC Journal, 2022, 73(8): 3647-3658.
[14]	Kun WANG, Hongbo SHI, Shuai TAN, Bing SONG, Yang TAO. Local time difference constrained neighborhood preserving embedding algorithm for fault detection [J]. CIESC Journal, 2022, 73(7): 3109-3119.
[15]	Xin ZHANG, Li ZHOU, Shihui WANG, Xu JI, Kexin BI. Integrated optimization of refinery hydrogen networks with crude oil properties fluctuations [J]. CIESC Journal, 2022, 73(4): 1631-1646.